Hypothesis Testing
This hypothesis states that a treatment has no effect.
What is the null hypothesis (H₀)?
Rejecting a true null hypothesis is referred to as this type of error.
What is a Type I error?
The t statistic is used instead of z when this population value is unknown.
What is the population standard deviation (σ)?
This design compares two completely different groups of participants.
What is an independent-measures (between-subjects) design?
This design uses the same participants in two time-separated conditions.
What is a repeated-measures (within-subjects) design?
This statistic tells us how large a treatment effect is, not just whether it exists.
What is effect size?
This probability value defines what counts as “very unlikely” in a hypothesis test.
What is alpha (α)?
Failing to detect a real treatment effect is called this.
What is a Type II error?
t distributions differ from z distributions because they are more ____.
What is spread out and variable ("fatter" tails)?
The null hypothesis for an independent samples t test states that these two values are equal.
What are μ₁ and μ₂ (the population means)?
The repeated-measures t test analyzes these values instead of raw scores.
What are difference scores (D)?
When conducting an independent-samples t test, this statistic must be calculated in order to compute the estimated standard error.
What is pooled variance?
A "statistically significant" result typically means you can do this regarding the null hypothesis.
What is reject the null?
This symbol represents the probability of a Type I error.
What is alpha (α)?
Degrees of freedom for a one-sample t test are calculated this way.
What is sample size, minus 1 (df = n − 1)?
This assumption requires both populations to have similar variances.
What is homogeneity of variance?
The alternative hypothesis for a paired-samples t test states that this value would not equal zero.
What is μᴰ (the population mean difference)?
This assumption requires that each observation in a sample be unrelated to the others.
What is independence of observations?
When interpreting the results of a hypothesis test, we are describing the results in comparison to this.
What is the null hypothesis?
This term refers to the probability that a statistical test can detect a real effect.
What is statistical power?
This value estimates how much sample means vary around the population mean.
What is the estimated standard error (sₘ)?
This test checks homogeneity of variance when sample sizes are unequal.
What is Levene’s test?
Order effects are best controlled for using this technique.
What is counterbalancing?
Compared to two-tailed tests, a one-tailed test is more likely to result in this type of error.
What is Type I error?
This statistic represents the probability of obtaining test results at least as extreme as what was actually observed.
What is the p-value?
This list represents the three factors which can increase statistical power.
What are larger sample size, larger effect size, and using a one-tailed test?
As sample size increases, the t distribution becomes more like this distribution.
What is the normal (z) distribution?
If a confidence interval for μ₁ − μ₂ includes zero, the result is ____.
What is not statistically significant?
When comparing independent-samples t tests to related-samples t tests, these are the advantages of a paired design.
What are lower variability, lower sample size required, and reduced effects of individual differences/demand characteristics.
When reporting the results of a t test, these statistics should be included.
What are the t value along with degrees of freedom, the p-value, and descriptives (M and SD).